Prahl Alison, Schuele C Melanie
Department of Communication Sciences and Disorders, Baylor University, Waco, TX, United States.
Department of Hearing and Speech Sciences, Vanderbilt University Medical Center, Nashville, TN, United States.
Front Psychol. 2022 Aug 12;13:905273. doi: 10.3389/fpsyg.2022.905273. eCollection 2022.
Obtaining valid assessments of language and literacy skills in children with Down syndrome (DS) presents a challenge as there is a paucity of information about the psychometrics of measures that are commonly used to measure listening and reading comprehension in this population. Evaluating the construct validity of measures that employ different methods is essential to ascertain the optimal method of assessment in individuals with DS and with typical developmental histories (TD). This pilot study aimed to evaluate the construct validity of four parallel measures of listening and reading comprehension. Participants included 19 individuals with DS ( = 17 years, 3 months; = 3 years, 6 months) and 19 word-level reading-matched children with TD ( = 7 years, 2 months; = 7 months). Participants completed norm-referenced assessments for four parallel measures of listening and reading comprehension. The four measurement methods were: (1) non-verbal response, (2) cloze procedure, (3) passage-level with close-ended questions, and (4) passage-level with open-ended questions. Participants completed additional assessments (e.g., cognition, language, and speech) for descriptive purposes. Construct validity was assessed using the Multitrait-Multimethod Matrix, a correlation matrix arranged to facilitate the assessment and interpretation of construct validity of measures across various formats. For both study groups, we observed strong evidence of construct validity for three out of four measurement methods. Results using the multimethod perspective also indicated that the listening and reading comprehension constructs were not separable. The findings from this pilot study represent a first step toward determining optimal methods of listening and reading comprehension assessment for individuals with DS. Additionally, these results can inform outcome measure selection in future language and literacy research with children with DS.
对唐氏综合征(DS)患儿的语言和读写能力进行有效评估面临挑战,因为关于该人群中常用的听力和阅读理解测量方法的心理测量学信息匮乏。评估采用不同方法的测量工具的结构效度,对于确定DS个体和典型发育史(TD)个体的最佳评估方法至关重要。这项初步研究旨在评估四种平行的听力和阅读理解测量工具的结构效度。参与者包括19名DS个体(平均年龄 = 17岁3个月;标准差 = 3岁6个月)和19名与单词水平阅读匹配的TD儿童(平均年龄 = 7岁2个月;标准差 = 7个月)。参与者完成了四种平行的听力和阅读理解测量工具的常模参照评估。这四种测量方法分别是:(1)非言语反应,(2)完形填空程序,(3)带有封闭式问题的篇章水平,以及(4)带有开放式问题的篇章水平。参与者还完成了其他评估(如认知、语言和言语)以作描述之用。使用多特质-多方法矩阵评估结构效度,该矩阵是一种相关矩阵,旨在便于评估和解释各种形式测量工具的结构效度。对于两个研究组,我们观察到四种测量方法中有三种具有强有力的结构效度证据。从多方法角度得出的结果还表明,听力和阅读理解结构不可分离。这项初步研究的结果是朝着确定DS个体听力和阅读理解评估的最佳方法迈出的第一步。此外,这些结果可为未来DS儿童语言和读写能力研究中的结果测量选择提供参考。